LAGLIDADG homing endonucleases ('LHEs', also termed 'meganucleases'), zinc finger nucleases ('ZFNs') TAL effector nucleases '(TALENs') and CRISPR nucleases are DNA cleavage systems that are used for genome engineering and corrective gene therapy. These systems display specificity profiles that correspond to varying amounts of off-target activity in the human genome. To address this issue, we have (1) developed methods for LHE engineering that is appropriate for academic laboratories (PNAS 2011); (2) determined the structure and recognition mechanism of a TAL effector (Science 2012); and (3) created and characterized a hyperspecific gene targeting scaffold termed a 'MegaTAL' endonuclease, that exploits the combined properties and mechanisms of TAL effectors and meganucleases. Aim 1: Create engineered MegaTALs to target two individual human disease-associated loci, then correlate their in vitro properties to their activities in transfected human cell lines. We have generated meganucleases that nick or cleave DNA target sites associated with two significant human genetic disorders (hemoglobinopathies and cystic fibrosis). The first of these therapeutic targets requires ex vivo disruption of a silencing regionin the -globin locus in patient-derived hematopoietic cell lines, whereas the second application requires in vivo targeted gene correction in the lung epithelium. We hypothesize that the MegaTALs will display both increased cleavage activity (by localizing the endonuclease to the target) and exceptional specificity for that same DNA sequence. We will test the hypothesis by characterizing: (1) the effect of the TAL anchor on gene conversion levels; and (2) the effect of the TAL anchor on genome-wide cleavage profiles and off-target activities. Aim 2: Augment the MegaTAL scaffold with tailored nucleolytic activities and molecular recruitment domains that can reduce undesired repair outcomes and/or enhance gene conversion activity. We have developed strategies both to diversify the nucleolytic activity of the LHE (to generate site specific nickases and cleavases for the same targets) and to recruit corrective DNA templates and/or recombination machinery to the target by the meganuclease. We hypothesize that these constructs can be exploited to further reduce or eliminate undesirable off-target activity, mutagenesis and toxicity, and will test that hypothesis using a panel of cellular assays for DNA repair, fidelity and viability as described in the proposal.